That edit is the part I wish had been your thesis. (Because now I feel like I have to write the following:)
There _is_ a unifying theme to the trends TechCrunch is writing about, but it's not what TechCrunch says it is. TechCrunch: "the establishment is flocking to the tech industry."
That's just the eternal september, ad infinitum. Hate being pushed out? So did the people _you_ pushed out, when you were annoying and first found IRC and you and your peers flooded all the old cool dives, driving hackers ever deeper underground.
The easy ways out don't work when you're faced with filtering out annoying newcomers. Some are fundametally wrong (racism is essentially a quick band-aid reaction to newcomers). Using money as a filter (country clubs) doesn't work either--there's always that annoying newly-minted millionaire who wants to break into your circle. Legislation to suppress young upstart competition even fails for the same reason: eventually your industry becomes so totally ossified that the whole thing can be toppled (US automakers).
Fundamentally, anybody who doesn't get Arrow's Impossibility Theorem is going to fail. "When voters have three or more distinct options, no rank order voting system can convert the ranked preferences of individuals into a community-wide (complete and transitive) ranking while also meeting a pre-specified set of criteria." I would propose strengthening it to: any pre-specified set of criteria, at all, ever. In other words, you just can't have all three:
1. Purely democratic or meritocratic or whatever
2. Total transparency (of rules and laws)
3. Adaptation of the rules and laws
So you get this never-ending cycle: the hackers get together and make something really cool. They are able to work together. Then comes the flood of newbies. After things get overloaded, anybody actually contributing to the original project either sells out or goes deeper underground. The project coasts on autopilot until something breaks, then it crashes and burns.
It doesn't matter who the participants are, from simple rule-based sims up to human-level intelligence, the outcome is the same.
The exceptions, I think, prove the rule:
1. BSD and the Linux kernel -- essentially ruled by benevolent dictators.
2. Representative democracies -- they can work. I'm deliberately refusing to think about why here, since that's politics.
3. The internet. Not IP or the OSI model or the DNS system, but the idea that anyone, anywhere can conceive of an idea, and then communicate it nearly instantly. This includes a worldwide language (or tools to automatically translate), global comms and the hacks to circumvent state-sponsored net terrorism, the unwritten code of conduct that most netizens are guided by, the ability to search and explore, and on and on.
I think any good example would have some element of a social hack in it. The social hacks are the most interesting.